jointly, and where there is also unpaid but essential household production work to be
done. The trade-oVis therefore not simply one between net income and leisure, but
between consumption goods bought in the market and having more time for
household activities, and also between the incomes and non-working time of hus-
band and wife. Moreover, lifetime considerations may be important, as people may
work hard during their prime-age years to provide for their (early) retirement.
Thus, economic theory, certainly when some model assumptions are relaxed,
cannot provide a clear-cut answer as regards the direction of the eVect of real
world tax-and-transfer systems, and moreover, theory is silent on themagnitudeof
the eVects, which is as important as the direction. Empirical studies only can provide
useful answers. There are several approaches in this domain. One is to use real-world
socioeconomic experiments, of which the best-known example is probably the New
Jersey negative income tax experiment (Pechman and Timpane 1975 ). The broad
conclusion from this and other similar experiments was that there was a noticeable
but not massive reduction in work eVort (Atkinson 1993 a, 43 ). Yet, although the
evidence produced by such experiments is unique, it cannot be regarded as conclu-
sive, for the reasons set out in Section 2. Other studies have followed the before-after
method, or the modeling approach outlined in Section 2.
Atkinson ( 1993 b, 297 ), reviewing a number of such studies, concludes that, overall,
‘‘a number of the eVects that have been identiWed are relatively small in size,’’ and
‘‘there are relatively few situations in which a disincentive eVect has been clearly
established.’’ There is evidence that taxation causes married women to work less, but
little evidence of a negative response by prime-age male workers. There is also little
clear evidence thatbeneWtsrepresent a major discouragement to take up work. One
reason for this is that, though the tax-and-transfer system in many countries creates a
poverty trap, this may aVect relatively few people. Also, transfers may have a positive
impact (the so-called entitlement eVect), as people keep working or looking for work
in order to become or remain eligible for beneWts.
Another group for which tax-and-transfer arrangements may have an important
eVect on labor market participation (apart from married women) is men aged 50 – 64.
In many countries participation rates for this group have fallen drastically during the
last four decades. Gruber and Wise ( 1998 ) show that, across a number of OECD
countries, labor force participation of older persons is strongly related to the implicit
social security tax on work. This implicit tax arises because in many countries,
staying on for one more year in the labor force for older persons implies a reduction
in the present discounted value of total pension beneWts during the remaining
lifetime. In some cases, this reduction is even larger than the net wages earned during
the extra period in work! The ‘‘tax force to retire’’ is especially strong in Italy,
Belgium, the Netherlands, Germany, and France. However, as Gruber and Wise
note, in some countries (e.g. Belgium) the reduction in labor market participation
of older persons was not an unwanted side product; rather, encouraging older
workers to leave the labor force was an explicit goal, with a view to easing labor
market tension and reducing unemployment among younger workers.
policy impact 311